The proposed research focuses on the behavioral and neurologic factors influencing spoken language (sound-to-word) learning, specifically the learning of non-native lexical tones and consonants in word identification in adulthood. Our behavioral-neural approach will enable the pursuit of our long-term goal of seeking the most desirable learning outcome mediated by behavioral training. The ten planned experiments in this grant application employ techniques that include behavioral training, neuroanatomic characterizations of learners of different learning abilities, and measurements of neurophysiologic (cerebral hemodynamic) changes associated with learning. The specific aims are to: 1) Compare the efficacy of non-native lexical tone and consonant training programs that emphasize low- versus high-stimulus variability in the training stimuli, and to compare neurophysiologic responses (measured by event-related fMRI) associated with the two training programs; 2) Investigate whether poor lexical learning can be remediated by additional training that focuses on phonetic (non-lexical pitch and consonant) learning and to examine neurophysiologic changes associated with such remediation; and 3) Characterize neuroanatomic differences between learners of different abilities. Guided by our newly proposed model of spoken language processing and learning called the Integrated Spoken Language Acquisition Network, we hypothesize the following: 1) Training with high-stimulus variability is the most efficacious, which is indicated by increased brain responses in auditory association cortex and the recruitment of the parietal lobe; 2) poor word learning resulting from a lack of phonetic representation and can be remediated by first training to establish such representations before word learning; and 3) poor learners can be identified by a relatively smaller left Heschl's Gyrus (driven by grey or white matter depending on the acoustic feature); however, successful learning can still be achieved with the appropriate remedial training. By understanding the efficacy of different training programs and how they benefit individuals of different learning profiles, as well as by identifying the neural characteristics of the different learners before and after training, we move a step closer to being able to place learners into training programs that are likely to be the most cost- effective and will likely lead to the most desirable learning outcomes. In an increasingly multi-lingual/multi-cultural world, many people, including adults, are interested in learning a foreign language even though it is difficult for them to do so. Learning a foreign language requires resources in our brain and changes occur as a result of learning. Our proposed research seeks to understand the changes that occur in the brain as adults learn to use foreign sounds in words, what effects different training methods have on the brain, and why certain individuals can learn more successfully (e.g., whether they have different brain organization before and after training).